Entity map
This is the machine-readable knowledge map for Hartzer.it.com, published to the EntityMap v1.0 specification. It describes the subjects this reference covers — the AI search surfaces, the techniques publishers are told to adopt, and the guides that connect them — with extractive passages taken from the pages themselves and attributed to them.
Every passage below is lifted from a real page rather than written for this file. Every relation points at an entity defined in this same file. External identifiers appear only where the grounding pass verified them; an entity with no verified identifier carries none rather than a guess.
View the EntityMap JSON → 43 entities · EntityMap v1.0
AI Overviews
The best-documented AI search surface, and the gap between what Google states and what practitioners sell.
Same as: https://en.wikipedia.org/wiki/AI_Overviews
Relations:
- PART_OF → AI Search Surfaces
e_grp_surface
AI Overviews is a feature of Google Search that places a generated summary at or near the top of the results page for some queries. The summary is written by a Gemini model working alongside Google's ranking systems and the Knowledge Graph, and it carries links out to pages that Google says support the statements it makes. Nobody opts in per query.
Google AI Overviews: What the Evidence Shows | Hartzer.it.com — published by Hartzer.it.com
Documented. Google documents the model, the eligibility rule and the controls; which eligible pages get cited it has never documented.
Google AI Overviews: What the Evidence Shows | Hartzer.it.com — published by Hartzer.it.com
AI Mode
Google's conversational search tab, the fan-out behind it, and why its citations barely overlap with AI Overviews.
Relations:
- PART_OF → AI Search Surfaces
e_grp_surface
AI Mode is a conversational search surface inside Google Search: rather than returning a ranked list of results, it returns a written answer assembled from several searches Google runs on the user's behalf, and the thread stays open for follow-up questions. Users reach it from a tab inside Search; it does not replace the ordinary results page. It is a different product from AI Overviews , which is a summary block on that ordinary page, and from the Gemini app, which sits outside Search altogether.
Google AI Mode: What Is Documented, What Is Not | Hartzer.it.com — published by Hartzer.it.com
Observed. Google names the mechanism and publishes none of its output; what is known about AI Mode citation comes from vendor measurement.
Google AI Mode: What Is Documented, What Is Not | Hartzer.it.com — published by Hartzer.it.com
ChatGPT search
The web retrieval layer inside ChatGPT: well documented as plumbing, entirely undocumented as a ranking system.
Same as: https://en.wikipedia.org/wiki/ChatGPT
Relations:
- PART_OF → AI Search Surfaces
e_grp_surface
ChatGPT search is the web retrieval layer inside OpenAI's ChatGPT assistant. When a prompt looks as though it needs current or external information, ChatGPT issues queries against search providers, reads a small set of the pages that come back, and writes an answer carrying inline links to the sources it used.
ChatGPT Search: Crawlers, Citations, Evidence | Hartzer.it.com — published by Hartzer.it.com
Unsupported. OpenAI documents its crawlers and its referral tag in detail, and publishes nothing at all about how sources are chosen.
ChatGPT Search: Crawlers, Citations, Evidence | Hartzer.it.com — published by Hartzer.it.com
Perplexity
The answer engine whose crawler documentation is exact, whose source selection is undocumented, and whose blocking advice is usually wrong.
Same as: https://en.wikipedia.org/wiki/Perplexity_AI
Relations:
- PART_OF → AI Search Surfaces
e_grp_surface
Perplexity AI is an answer engine: a search product whose default output is written prose rather than a list of results. A query triggers a live search of the web, the system reads the pages it retrieves, and it returns an answer carrying numbered inline citations, each one a link, with a strip of sources beside or above the text.
Perplexity: How It Cites, Crawls and Pays | Hartzer.it.com — published by Hartzer.it.com
Unsupported. Perplexity's own documentation contradicts the standard advice for blocking it, and says nothing about how sources are chosen.
Perplexity: How It Cites, Crawls and Pays | Hartzer.it.com — published by Hartzer.it.com
Gemini
The answer surface Google documents least, and the one every optimization claim is guessing about.
Relations:
- PART_OF → AI Search Surfaces
e_grp_surface
Gemini is Google's family of generative AI models and the consumer app built on them; treated as a search surface, it is a place where a person asks a question and receives a written answer that may, but need not, carry links to web pages. Those links come from grounding — at answer time the model issues queries against the Google Search index, retrieves current pages, and attaches them to its response as related sources, which is what separates an answer reflecting the live web from one composed entirely out of model weights.
Google Gemini as a Search and Answer Surface | Hartzer.it.com — published by Hartzer.it.com
Unsupported. Google publishes no eligibility rule, no selection criteria and no reporting for Gemini, so every lever sold for it is inferred.
Google Gemini as a Search and Answer Surface | Hartzer.it.com — published by Hartzer.it.com
Microsoft Copilot
Bing-grounded generative answers, page-level publisher controls, and the only citation-impression reporting in the field.
Same as: https://en.wikipedia.org/wiki/Microsoft_Copilot
Relations:
- PART_OF → AI Search Surfaces
e_grp_surface
Microsoft Copilot is Microsoft's consumer AI assistant, and Copilot Search is the generative answer layer inside Bing. Both are grounded on the Bing index: Bing performs the crawl and the retrieval, and a language model writes an answer over the retrieved passages with inline links back to the sources. The name has moved twice. It launched as Bing Chat on 7 February 2023, became Copilot on 15 November 2023, and the brand now covers a consumer assistant, Copilot inside Windows and Microsoft 365, and Copilot Studio for building enterprise agents.
Microsoft Copilot: Citations and Bing Reporting | Hartzer.it.com — published by Hartzer.it.com
Observed. Microsoft reports citations to publishers better than any other operator, and still publishes no criteria for what gets cited.
Microsoft Copilot: Citations and Bing Reporting | Hartzer.it.com — published by Hartzer.it.com
AI assistants and smaller engines
The assistants and engines too small to warrant a page each, compared on the only thing that separates them: what they document.
Relations:
- PART_OF → AI Search Surfaces
e_grp_surface
An AI assistant, in the sense this page uses, is a conversational product that answers a question by fetching live web pages and, usually, showing the reader where the answer came from. This page covers the ones that are not ChatGPT, not Google and not Microsoft: Claude from Anthropic, Meta AI, DuckDuckGo's AI-assisted answers, Brave Search, Grok from xAI, and Le Chat from Mistral. They are grouped because none of them individually justifies a page, and pretending otherwise would misrepresent the evidence.
AI Assistants: Claude, Meta AI, DuckDuckGo, Brave — published by Hartzer.it.com
Unsupported. Not one of these operators documents how it selects sources, and not one documents how a publisher should measure it.
AI Assistants: Claude, Meta AI, DuckDuckGo, Brave — published by Hartzer.it.com
Generative Engine Optimization
The founding paper, the misquoted 40%, and the replication attempt that found three significant results out of fifty-four.
Same as: https://en.wikipedia.org/wiki/Generative_engine_optimization
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
Generative engine optimization (GEO) is the practice of changing a web page, or a body of content, so that a generative AI system — one that answers a question in prose and attaches citations, such as Google AI Overviews, Google AI Mode, ChatGPT search, Perplexity or Microsoft Copilot — becomes more likely to retrieve it, cite it, or repeat its claims. That is the definition the term was born with, and it is narrower than the one sold.
Generative Engine Optimization: The Evidence | Hartzer.it.com — published by Hartzer.it.com
Unsupported. The term and the paper behind it are real; the content tactics sold under the name fail almost every independent test run on them.
Generative Engine Optimization: The Evidence | Hartzer.it.com — published by Hartzer.it.com
Answer Engine Optimization
The term predates generative AI, the tactic it was built on has been withdrawn, and no operator documents it under that name.
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
Answer engine optimization (AEO) is the practice of structuring content so that a system returning a single direct answer — rather than a list of blue links — uses that content as the answer. An answer engine, in this sense, is any product that collapses a result set into one response: a voice assistant reading a result aloud, a featured snippet, or in 2026 an AI assistant writing a paragraph with citations attached. The term predates generative AI by roughly five years.
Answer Engine Optimization: Is AEO Distinct? | Hartzer.it.com — published by Hartzer.it.com
Unsupported. No operator documents AEO, Wikipedia redirects it to GEO, and a vendor selling it says in print the two describe one job.
Answer Engine Optimization: Is AEO Distinct? | Hartzer.it.com — published by Hartzer.it.com
GEO vs AEO vs SEO
Where each acronym came from, who gained by pushing it, what actually changed about the surface, and the split verdict on whether classic SEO is enough.
Same as: https://en.wikipedia.org/wiki/Search_engine_optimization
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
Search engine optimization (SEO) is the long-established practice of making a page findable and rankable in a search index; it has been in documented use since 1997 and predates Google. Generative engine optimization (GEO) comes from a 2023 academic preprint and names the practice of influencing what a generative model says and cites. Answer engine optimization (AEO) is older than GEO, in circulation by early 2018 for voice search and featured snippets, and has since been re-pointed at the same AI surfaces GEO addresses.
GEO vs SEO vs AEO: What Actually Differs | Hartzer.it.com — published by Hartzer.it.com
Unsupported. Sold as three disciplines, the split has no evidence: no operator documents anything that answers to one label and not the others.
GEO vs SEO vs AEO: What Actually Differs | Hartzer.it.com — published by Hartzer.it.com
AI Citations and Source Selection
The eligibility rule is documented and the selection function is not, which is why four studies of the same question produced four different numbers.
Same as: https://en.wikipedia.org/wiki/Citation
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
Source selection is not one step. An AI answer surface decides which pages to retrieve , which of those to actually use when composing the answer, and which to show the reader as a citation — a visible attribution, usually a clickable link, beneath or beside the generated text. Those three sets are not the same set. A page can be retrieved and never used, used and never cited, or cited while contributing almost nothing to the sentence it sits under.
AI Citations and Source Selection | Hartzer.it.com — published by Hartzer.it.com
Observed. Eligibility for citation is documented; how sources are chosen and credited is not, and the published figures contradict each other.
AI Citations and Source Selection | Hartzer.it.com — published by Hartzer.it.com
Brand Mentions in AI Search
Two different mechanisms get sold as one. One follows from documented retrieval behavior; the other is an inference about training data that nobody outside the labs can test.
Same as: https://en.wikipedia.org/wiki/Brand_awareness
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
A brand mention is any occurrence of a brand's name — the name a company trades under, and the associations attached to it — in text on a page that brand does not control. An unlinked mention is one with no hyperlink pointing back: the name appears, nothing points anywhere. Under classic SEO an unlinked mention was a curiosity, something to convert into a link. Since 2025 it has been sold as the primary input to AI search visibility, on the reasoning that language models consume text rather than link graphs.
Brand Mentions in AI Search | Hartzer.it.com — published by Hartzer.it.com
Unsupported. Mentions and AI visibility move together in every study; no published experiment shows that adding mentions moves anything.
Brand Mentions in AI Search | Hartzer.it.com — published by Hartzer.it.com
Schema Markup for AI Search
The correlation between schema and AI citations is real, the causal evidence is null, and Google has written down which half matters.
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
Structured data is machine-readable metadata added to a web page describing what the page is about — a product, a person, an organization, an event — using a shared vocabulary. On the web that vocabulary is almost always schema.org , launched on 2 June 2011 by Google, Microsoft and Yahoo, and the dominant serialization is JSON-LD (JavaScript Object Notation for Linked Data), a block of JSON placed inside a script tag on the page. Its historical payoff was concrete and visible.
Schema Markup for AI Search: The Evidence | Hartzer.it.com — published by Hartzer.it.com
Unsupported. Google states no special markup is needed for its AI features, and the one controlled test found no citation uplift anywhere.
Schema Markup for AI Search: The Evidence | Hartzer.it.com — published by Hartzer.it.com
FAQ Markup and Answer Extraction
Google ended FAQ rich results on 7 May 2026 with no blog post. The editorial pattern survived the markup, and they were never the same thing.
Same as: https://en.wikipedia.org/wiki/Question_answering
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
They have different mechanisms and, since May 2026, very different fates. Keeping them apart is most of the work on this subject. FAQPage structured data is schema.org markup declaring that a page contains a list of questions and their answers, each question a Question with an acceptedAnswer . Between 2018 and 2023 it made a page eligible for an FAQ rich result — an expandable accordion attached to the page's listing in Google's results, which visibly enlarged the listing and was widely credited with lifting click-through.
FAQ Markup and Answer Extraction in AI Search | Hartzer.it.com — published by Hartzer.it.com
Unsupported. FAQ rich results were removed from Google Search entirely on 7 May 2026; the markup now produces no search appearance at all.
FAQ Markup and Answer Extraction in AI Search | Hartzer.it.com — published by Hartzer.it.com
Content Chunking and Passage Retrieval
What chunking and passage retrieval actually are, why passage ranking is a different mechanism, and where the recommended section lengths came from.
Same as: https://en.wikipedia.org/wiki/Information_retrieval
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
Chunking is the step in a retrieval pipeline where a document is cut into smaller units before it is indexed, so the system can return a relevant fragment instead of a whole file. Passage retrieval is the read side of the same idea: matching a query against those fragments and returning the ones that answer it. Both belong to information retrieval , the discipline concerned with finding material that satisfies a need from within a large collection.
Content Chunking and Passage Retrieval in AI Search — published by Hartzer.it.com
Observed. The retrievable unit really is the passage, but no operator publishes a chunk size and Google says not to fragment your content.
Content Chunking and Passage Retrieval in AI Search — published by Hartzer.it.com
Entity Grounding and the Knowledge Graph
Google documents what the Knowledge Graph is and what Organization markup does. Nobody has published a study connecting either to being cited in an AI answer.
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
An entity is a thing with an identity — a person, a company, a place, a product, an event — as distinct from the string of characters used to name it. Google's framing when it launched the Knowledge Graph in 2012 was "things, not strings." Entity grounding , also called entity linking or entity resolution, is the step in which a system reads a name in text and decides which known thing it refers to: that this "Apple" is the company rather than the fruit, and that the surname in one document belongs to the same person as the surname in another.
Entity Grounding and the Knowledge Graph | Hartzer.it.com — published by Hartzer.it.com
Unsupported. The Knowledge Graph is documented and real; that being in it makes an AI answer cite you has never been measured, either way.
Entity Grounding and the Knowledge Graph | Hartzer.it.com — published by Hartzer.it.com
E-E-A-T Signals in AI Search
E-E-A-T is the most documented concept in search and the least documented in relation to AI answers. What fills that gap is the old checklist, relabeled.
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
E-E-A-T stands for Experience, Expertise, Authoritativeness and Trust. It comes from Google's Search Quality Rater Guidelines , the manual issued to the external human contractors Google pays to rate sample search results. Raters apply it when judging Page Quality. It was E-A-T from 2014; Google added the leading Experience — first-hand or life experience with the subject — on 15 December 2022. What it is not is a score.
E-E-A-T in AI Search: What Google Documents | Hartzer.it.com — published by Hartzer.it.com
Unsupported. No AI operator documents author or credential signals, and Google's May 2026 AI guide never uses the term E-E-A-T.
E-E-A-T in AI Search: What Google Documents | Hartzer.it.com — published by Hartzer.it.com
AI Crawler Directives and llms.txt
The tokens that cost you AI visibility if you block them, the ones that only touch training, and the server-log study that settles llms.txt.
Same as: https://en.wikipedia.org/wiki/Web_crawler
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
AI crawler directives are entries in a site's robots.txt naming the user-agent tokens that AI companies use, and allowing or disallowing them. The mechanism is the ordinary Robots Exclusion Protocol, standardized as RFC 9309 in September 2022 — the same file and the same syntax that has governed every web crawler , the automated program that fetches pages to build an index, since the 1990s.
AI Crawler Directives and llms.txt | Hartzer.it.com — published by Hartzer.it.com
Unsupported. Crawler tokens work as their operators document; llms.txt is read by almost nobody — 97% of files got zero requests in May 2026.
AI Crawler Directives and llms.txt | Hartzer.it.com — published by Hartzer.it.com
Retrieval-Augmented Generation
The retrieve-then-generate pipeline explained for search practitioners, with the parts the operators document and the parts nobody does.
Same as: https://en.wikipedia.org/wiki/Retrieval-augmented_generation
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
Retrieval-augmented generation (RAG) is the technique of answering a question by first retrieving relevant text from an external source, then handing that text to a language model as part of its prompt, so the model writes its answer from documents rather than only from what it absorbed during training. Generation, here, means the model composing prose token by token; retrieval means fetching candidate text before it starts.
Retrieval-Augmented Generation for AI Search | Hartzer.it.com — published by Hartzer.it.com
Documented. Google, Microsoft and OpenAI all document the retrieve-then-generate loop; not one of them documents how sources are chosen inside it.
Retrieval-Augmented Generation for AI Search | Hartzer.it.com — published by Hartzer.it.com
Query Fan-Out
The mechanism is one of the few things Google has volunteered about how AI answers are built. The variant taxonomies and iteration counts around it are patent readings, not product documentation.
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
Query fan-out is the practice of an AI search surface taking one question from a user, breaking it into several related sub-queries — also called synthetic queries or query variants — running those searches at the same time against one or more indexes, and composing a single answer from the combined results. The user types one thing; the system searches for many things. The term is Google's own, and it is one of very few mechanical details Google has volunteered about how its AI answers are assembled.
Query Fan-Out in AI Search | Hartzer.it.com — published by Hartzer.it.com
Documented. Google names and documents the technique, and has never published how many sub-queries it runs or how they are generated.
Query Fan-Out in AI Search | Hartzer.it.com — published by Hartzer.it.com
AI Visibility Tracking
The practice is real and partly documented by Google and Microsoft. The single visibility score built on top of it is a vendor construction with an error bar wider than the changes it reports.
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
AI visibility tracking is the practice of measuring whether, how often and in what form a brand or a website appears inside answers produced by generative search surfaces — Google AI Overviews and AI Mode, ChatGPT, Perplexity, Microsoft Copilot and Gemini among them. It is a measurement discipline rather than a product, and it exists because the answer a person reads is assembled at request time rather than looked up in a fixed index.
What AI Visibility Tracking Can Measure | Hartzer.it.com — published by Hartzer.it.com
Observed. Two platforms document fragments of it; every composite visibility score on the market is a vendor construction no platform confirms.
What AI Visibility Tracking Can Measure | Hartzer.it.com — published by Hartzer.it.com
AI Overview Rank Tracking
Presence and inclusion can be sampled honestly. Citation position cannot: Google assigns the whole block one position, and 45% of citations change between generations.
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
AI Overview rank tracking is the attempt to apply the logic of keyword rank tracking — a keyword list, a repeated measurement, a position number, a trend line — to Google's AI Overviews, the generated summary block that appears above or among the classic results on a Google search page. In practice it means answering three questions per keyword on a schedule. Presence. Does this keyword trigger an AI Overview at all? Inclusion. Is my domain cited as a source inside it? Placement.
AI Overview Rank Tracking: What Breaks | Hartzer.it.com — published by Hartzer.it.com
Unsupported. Presence and inclusion can be sampled; the citation rank trackers report is a construct Google's own documentation denies.
AI Overview Rank Tracking: What Breaks | Hartzer.it.com — published by Hartzer.it.com
AI Search Attribution and Analytics
Only OpenAI documents a tagging convention. Every other AI referrer is a string nobody promised to keep, and Google's AI surfaces arrive looking exactly like organic search.
Relations:
- PART_OF → Techniques and Signals
e_grp_technique
AI search attribution is the problem of establishing, from your own analytics and server logs, that a visitor arrived because of an AI answer — which surface sent them, for what, and what they did next. It is a strictly narrower question than AI visibility. Visibility asks whether you appeared; attribution asks whether appearing produced anything. The plumbing is ordinary and decades old.
AI Search Attribution and Analytics | Hartzer.it.com — published by Hartzer.it.com
Observed. One operator documents a tag; every other AI referrer is a practitioner-observed string nobody has promised to keep working.
AI Search Attribution and Analytics | Hartzer.it.com — published by Hartzer.it.com
Understanding AI search
What AI search is, how a generated answer is assembled, and how to read the studies that claim to measure it.
Relations:
- PART_OF → Understanding AI Search
e_grp_understanding - RELATES_TO → AI Overviews
e_google_ai_overviews - RELATES_TO → AI Mode
e_google_ai_mode - RELATES_TO → ChatGPT search
e_chatgpt_search - RELATES_TO → AI Citations and Source Selection
e_ai_citations_and_source_selection
AI search is the class of products that answer a question in generated prose, assembled at the moment of asking from pages retrieved off the live web, with links to some of those pages attached. Three parts carry the weight. The text is written by generative AI — a language model producing sentences — rather than selected from a page; the sources are fetched in response to the question rather than recalled from training; and the attribution, where it exists, is a display decision made after the prose is written, not a ranking published in advance.
What AI Search Is and How It Works | Hartzer.it.com — published by Hartzer.it.com
Optimizing for AI search
The documented requirements for appearing in AI answers, and the widely sold tactics that failed replication.
Same as: https://en.wikipedia.org/wiki/Search_engine_optimization
Relations:
- PART_OF → Optimizing for AI Search
e_grp_optimizing - RELATES_TO → Generative Engine Optimization
e_generative_engine_optimization - RELATES_TO → Answer Engine Optimization
e_answer_engine_optimization - RELATES_TO → GEO vs AEO vs SEO
e_geo_vs_seo - RELATES_TO → Schema Markup for AI Search
e_schema_markup_for_ai_search
Google's guide to optimizing for generative AI features carries a section headed Mythbusting generative AI search: what you don't need to do — an explicit list of the things Google says are unnecessary.
Optimizing for AI Search: What the Evidence Says | Hartzer.it.com — published by Hartzer.it.com
Measuring AI search visibility
What the platform reports contain, why one measurement is one draw from a distribution, and what a defensible AI visibility report discloses.
Relations:
- PART_OF → Measuring AI Search Visibility
e_grp_measuring - RELATES_TO → AI Visibility Tracking
e_ai_visibility_tracking - RELATES_TO → AI Search Attribution and Analytics
e_ai_search_attribution_and_analytics - RELATES_TO → AI Overview Rank Tracking
e_ai_overview_rank_tracking - RELATES_TO → Share of Voice in AI Search
e_share_of_voice_in_ai_search
AI search measurement is the practice of establishing whether, how often and in what form a site or a brand appears inside answers produced by generative surfaces — Google's AI Overviews and AI Mode, ChatGPT, Perplexity, Microsoft Copilot, Gemini. It is not one measurement but four, sold under a single label: Citation — a link to your URL appears as a source beneath or beside the answer. Mention — your brand is named in the answer prose, whether or not anything links to you. Impression — a platform counts your link as having been shown to a user.
How to Measure AI Search Visibility | Hartzer.it.com — published by Hartzer.it.com
How AI search engines choose what to cite
The retrieve, use and cite decisions pulled apart, with what each platform documents and what four contradictory overlap studies actually measured.
Same as: https://en.wikipedia.org/wiki/Information_retrieval
Relations:
- PART_OF → Understanding AI Search
e_grp_understanding - RELATES_TO → AI Citations and Source Selection
e_ai_citations_and_source_selection - RELATES_TO → Query Fan-Out
e_query_fan_out - RELATES_TO → AI Overviews
e_google_ai_overviews - RELATES_TO → AI Mode
e_google_ai_mode
Source selection in an AI answer is a problem in information retrieval — the discipline of finding documents that satisfy a query — with an extra step bolted on the end. A generative surface decides which pages to retrieve , which of those to actually use when composing prose, and which to show the reader as a citation . Those are three decisions, not one, and they do not produce the same set of pages. A page can be retrieved and never used, used and never cited, or cited while contributing almost nothing to the sentence it sits beside.
How AI Search Engines Choose What to Cite | Hartzer.it.com — published by Hartzer.it.com
What changed when AI Overviews launched
The rollout dated from SGE onward, each publisher click study named with its sample and funder, and the numbers that get confused with each other.
Same as: https://en.wikipedia.org/wiki/AI_Overviews
Relations:
- PART_OF → Understanding AI Search
e_grp_understanding - RELATES_TO → AI Overviews
e_google_ai_overviews - RELATES_TO → AI Mode
e_google_ai_mode - RELATES_TO → AI Search Attribution and Analytics
e_ai_search_attribution_and_analytics - RELATES_TO → AI Citations and Source Selection
e_ai_citations_and_source_selection
Three boundaries matter more than everything else in this chronology, and they are the three most often got wrong. Search Generative Experience (SGE) was announced at Google I/O on 10 May 2023 as an opt-in Search Labs experiment, US English only. AI Overviews launched at I/O in May 2024 to all US users, with the SGE branding retired — the first time ordinary American searchers saw a generated answer without having chosen to.
What Changed When AI Overviews Launched | Hartzer.it.com — published by Hartzer.it.com
Why AI search traffic behaves differently
The three things called AI traffic, what the large panels agree and disagree on, and why the direct-traffic gap is a mechanism without a magnitude.
Same as: https://en.wikipedia.org/wiki/Web_traffic
Relations:
- PART_OF → Understanding AI Search
e_grp_understanding - RELATES_TO → AI Search Attribution and Analytics
e_ai_search_attribution_and_analytics - RELATES_TO → ChatGPT search
e_chatgpt_search - RELATES_TO → AI Overviews
e_google_ai_overviews - RELATES_TO → AI Mode
e_google_ai_mode
Web traffic — the sessions arriving at a site — splits into three distinct things once AI answers are involved, and almost every confused claim in this subject comes from quoting a figure about one of them as though it described another. AI referrals. Someone clicks a link inside an assistant's answer and arrives carrying a referrer. Measurable in ordinary analytics by referrer host, and in ChatGPT's case by a documented UTM parameter as well. Clicks inside Google.
Why AI Search Traffic Behaves Differently | Hartzer.it.com — published by Hartzer.it.com
How to Rank in AI Overviews
The honest answer to the most-asked question in AI search: there is no separate ranking system, and what genuinely follows from that is narrower than most advice admits.
Same as: https://en.wikipedia.org/wiki/AI_Overviews
Relations:
- PART_OF → Optimizing for AI Search
e_grp_optimizing - RELATES_TO → AI Overviews
e_google_ai_overviews - RELATES_TO → Query Fan-Out
e_query_fan_out - RELATES_TO → AI Citations and Source Selection
e_ai_citations_and_source_selection - RELATES_TO → E-E-A-T Signals in AI Search
e_eeat_signals_in_ai_search
There is no separate ranking system for AI Overviews . The generated summary at the top of a Google Search results page is written by a customized Gemini model working alongside the ranking systems Google already had, and it draws on the same index that produces the ordinary results underneath it.
How to Rank in Google AI Overviews | Hartzer.it.com — published by Hartzer.it.com
Structuring Content for Answer Extraction
The markup died in May 2026 with no blog post. The editorial pattern it was confused with was never the same thing and never depended on schema.
Relations:
- PART_OF → Optimizing for AI Search
e_grp_optimizing - RELATES_TO → FAQ Markup and Answer Extraction
e_faq_and_answer_extraction - RELATES_TO → Content Chunking and Passage Retrieval
e_content_chunking_and_passage_retrieval - RELATES_TO → Schema Markup for AI Search
e_schema_markup_for_ai_search - RELATES_TO → Retrieval-Augmented Generation
e_retrieval_augmented_generation
Answer extraction is what a search or AI system does to a page when it pulls a span of text out of it and presents that span as the answer, with or without a link back. Featured snippets, People Also Ask boxes and the sentence-level quoting inside a Google AI Overview are all answer extraction. Structuring content for it means writing so that some passage of your page answers a question completely on its own. Two other things are bundled under the same heading and have entirely different fates.
Structuring Content for Answer Extraction | Hartzer.it.com — published by Hartzer.it.com
Technical Requirements for AI Crawlers
Which user-agent token does what, which ones cost you visibility if you block them, and the one Google token that does not do what its name suggests.
Relations:
- PART_OF → Optimizing for AI Search
e_grp_optimizing - RELATES_TO → AI Crawler Directives and llms.txt
e_ai_crawler_directives_and_llms_txt - RELATES_TO → ChatGPT search
e_chatgpt_search - RELATES_TO → Perplexity
e_perplexity - RELATES_TO → AI Overviews
e_google_ai_overviews
AI crawler access is governed by the ordinary Robots Exclusion Protocol — the same robots.txt file, the same syntax, standardized as RFC 9309 in September 2022 — applied to a new set of user-agent tokens. The mechanism is not new. What is new is that the AI operators have split their crawling by purpose , so one company now publishes three or four tokens that do different jobs and carry different costs when blocked. Four limits apply to every rule you write, and three of them predate AI entirely. It is a crawl directive, not an index directive.
Technical Requirements for AI Crawlers | Hartzer.it.com — published by Hartzer.it.com
AI Search Visibility for Law Firms
Almost nothing legal-specific about AI search has been independently measured. What the evidence supports, and what the advertising rules already cover.
Relations:
- PART_OF → Optimizing for AI Search
e_grp_optimizing - RELATES_TO → AI Overviews
e_google_ai_overviews - RELATES_TO → Generative Engine Optimization
e_generative_engine_optimization - RELATES_TO → AI Citations and Source Selection
e_ai_citations_and_source_selection - RELATES_TO → Query Fan-Out
e_query_fan_out
When somebody with a legal problem asks an AI assistant a question — whether a landlord can evict them, how long they have to sue after a collision, whether they need a lawyer at all — the assistant answers in prose and usually cites a handful of sources. Occasionally it names firms. AI search for legal services is the question of which sources those are, whether a firm can influence the outcome, and what happens when an assistant says something about a firm that is not true. Start with the part that gets skipped.
AI Search Visibility for Law Firms | Hartzer.it.com — published by Hartzer.it.com
How to track AI Overview visibility
What can honestly be measured about an AI Overview citation, what Google has reported since June 2026, and how many samples a number needs before it means anything.
Relations:
- PART_OF → Measuring AI Search Visibility
e_grp_measuring - RELATES_TO → AI Overview Rank Tracking
e_ai_overview_rank_tracking - RELATES_TO → AI Overviews
e_google_ai_overviews - RELATES_TO → AI Mode
e_google_ai_mode - RELATES_TO → AI Visibility Tracking
e_ai_visibility_tracking
AI Overview visibility is not one measurement. Four different things are sold under the name, they come from different instruments, and they move independently of each other. Presence asks whether a query triggers an AI Overview at all. Inclusion asks whether your URL is among the sources cited inside it. Impression asks whether Google counted your link as shown to a user. Referral asks whether anyone clicked and arrived. Nothing forces those four to agree.
How to Track AI Overview Visibility | Hartzer.it.com — published by Hartzer.it.com
How to evaluate an AI visibility tracking tool
What a tool in this category can and cannot answer, why two honest tools disagree about the same brand, and the disclosures a number needs before it belongs in a report.
Relations:
- PART_OF → Measuring AI Search Visibility
e_grp_measuring - RELATES_TO → AI Visibility Tracking
e_ai_visibility_tracking - RELATES_TO → Share of Voice in AI Search
e_share_of_voice_in_ai_search - RELATES_TO → Brand Mentions in AI Search
e_brand_mentions_in_ai_search - RELATES_TO → Microsoft Copilot
e_microsoft_copilot
Tools in this category are bought as though they were interchangeable, and they are not. They answer different questions, and several of the questions buyers care about most have no tool behind them at all. Before comparing products, write down which of these you actually need answered. Are links to my pages being shown inside Google's generative answers? Google's own reporting answers this for your property, as a count rather than a sample.
Choosing AI Visibility Tracking Tools | Hartzer.it.com — published by Hartzer.it.com
What AI search analytics can and cannot tell you
A dated ledger of what Google Search Console, Bing Webmaster Tools, analytics and server logs each report about AI search, and the questions none of them answer.
Same as: https://en.wikipedia.org/wiki/Google_Search_Console
Relations:
- PART_OF → Measuring AI Search Visibility
e_grp_measuring - RELATES_TO → AI Search Attribution and Analytics
e_ai_search_attribution_and_analytics - RELATES_TO → AI Overviews
e_google_ai_overviews - RELATES_TO → AI Mode
e_google_ai_mode - RELATES_TO → Microsoft Copilot
e_microsoft_copilot
Google Search Console gained a generative AI performance report on 3 June 2026, with data beginning 18 May 2026. Google's definition of the metric is one sentence: impressions are how many times links to your site were shown to a user in a generative AI feature on Google Search. The report has four dimensions and no others — pages, countries, dates and devices — where pages groups data by the final URL linked after any redirects.
What AI Search Analytics Can and Cannot Tell You | Hartzer.it.com — published by Hartzer.it.com
Building an AI visibility report that survives scrutiny
What belongs in an AI visibility report, how to sample so that a change can be told apart from noise, and the comparisons to keep off the page.
Relations:
- PART_OF → Measuring AI Search Visibility
e_grp_measuring - RELATES_TO → AI Visibility Tracking
e_ai_visibility_tracking - RELATES_TO → Share of Voice in AI Search
e_share_of_voice_in_ai_search - RELATES_TO → AI Search Attribution and Analytics
e_ai_search_attribution_and_analytics - RELATES_TO → AI Overview Rank Tracking
e_ai_overview_rank_tracking
An AI visibility report exists to support a decision: whether to keep investing in a body of content, where the exposure is concentrated, and whether something changed that anyone should act on. It is not a scoreboard, and treating it as one is what produces the monthly ritual of explaining a number that moved for reasons nobody can name. Most of these reports fail for a reason that has nothing to do with data quality. They stack four different kinds of number in one chart — a count, a sample, a floor and a subset — and then draw a line through them.
Building an AI Visibility Report | Hartzer.it.com — published by Hartzer.it.com
AI Search Surfaces
The AI search surfaces covered by this reference: the products that answer a question in generated prose and attach links to some of their sources, rated on how much each operator documents about source selection.
Relations:
- INCLUDES → AI Overviews
e_google_ai_overviews - INCLUDES → AI Mode
e_google_ai_mode - INCLUDES → ChatGPT search
e_chatgpt_search - INCLUDES → Perplexity
e_perplexity - INCLUDES → Gemini
e_google_gemini - INCLUDES → Microsoft Copilot
e_microsoft_copilot - INCLUDES → AI assistants and smaller engines
e_ai_assistants_and_smaller_engines
The AI search surfaces covered by this reference: the products that answer a question in generated prose and attach links to some of their sources, rated on how much each operator documents about source selection.
AI Search Surfaces | Hartzer.it.com — published by Hartzer.it.com
Techniques and Signals
The techniques and signals covered by this reference: what a publisher does, or is told to do, to be cited by an AI answer, each rated as documented, observed, or unsupported by the available evidence.
Relations:
- INCLUDES → Generative Engine Optimization
e_generative_engine_optimization - INCLUDES → Answer Engine Optimization
e_answer_engine_optimization - INCLUDES → GEO vs AEO vs SEO
e_geo_vs_seo - INCLUDES → AI Citations and Source Selection
e_ai_citations_and_source_selection - INCLUDES → Brand Mentions in AI Search
e_brand_mentions_in_ai_search - INCLUDES → Schema Markup for AI Search
e_schema_markup_for_ai_search - INCLUDES → FAQ Markup and Answer Extraction
e_faq_and_answer_extraction - INCLUDES → Content Chunking and Passage Retrieval
e_content_chunking_and_passage_retrieval - INCLUDES → Entity Grounding and the Knowledge Graph
e_entity_grounding_and_the_knowledge_graph - INCLUDES → E-E-A-T Signals in AI Search
e_eeat_signals_in_ai_search - INCLUDES → AI Crawler Directives and llms.txt
e_ai_crawler_directives_and_llms_txt - INCLUDES → Retrieval-Augmented Generation
e_retrieval_augmented_generation - INCLUDES → Query Fan-Out
e_query_fan_out - INCLUDES → AI Visibility Tracking
e_ai_visibility_tracking - INCLUDES → AI Overview Rank Tracking
e_ai_overview_rank_tracking - INCLUDES → Share of Voice in AI Search
e_share_of_voice_in_ai_search - INCLUDES → AI Search Attribution and Analytics
e_ai_search_attribution_and_analytics
The techniques and signals covered by this reference: what a publisher does, or is told to do, to be cited by an AI answer, each rated as documented, observed, or unsupported by the available evidence.
Techniques and Signals | Hartzer.it.com — published by Hartzer.it.com
Understanding AI Search
What AI search is, how a generated answer is assembled, and how to read the studies that claim to measure it.
Relations:
- INCLUDES → How AI search engines choose what to cite
e_guide_how_ai_search_engines_choose_what_to_cite - INCLUDES → What changed when AI Overviews launched
e_guide_what_changed_when_ai_overviews_launched - INCLUDES → Why AI search traffic behaves differently
e_guide_why_ai_search_traffic_behaves_differently
What AI search is, how a generated answer is assembled, and how to read the studies that claim to measure it.
What AI Search Is and How It Works | Hartzer.it.com — published by Hartzer.it.com
Optimizing for AI Search
The documented requirements for appearing in AI answers, and the widely sold tactics that failed replication.
Relations:
- INCLUDES → How to Rank in AI Overviews
e_guide_how_to_rank_in_ai_overviews - INCLUDES → Structuring Content for Answer Extraction
e_guide_structuring_content_for_answer_extraction - INCLUDES → Technical Requirements for AI Crawlers
e_guide_technical_requirements_for_ai_crawlers - INCLUDES → AI Search Visibility for Law Firms
e_guide_ai_search_for_law_firms
The documented requirements for appearing in AI answers, and the widely sold tactics that failed replication.
Optimizing for AI Search: What the Evidence Says | Hartzer.it.com — published by Hartzer.it.com
Measuring AI Search Visibility
What the platform reports contain, why one measurement is one draw from a distribution, and what a defensible AI visibility report discloses.
Relations:
- INCLUDES → How to track AI Overview visibility
e_guide_how_to_track_ai_overview_visibility - INCLUDES → How to evaluate an AI visibility tracking tool
e_guide_choosing_ai_visibility_tracking_tools - INCLUDES → What AI search analytics can and cannot tell you
e_guide_what_ai_search_analytics_can_and_cannot_tell_you - INCLUDES → Building an AI visibility report that survives scrutiny
e_guide_building_an_ai_visibility_report
What the platform reports contain, why one measurement is one draw from a distribution, and what a defensible AI visibility report discloses.
How to Measure AI Search Visibility | Hartzer.it.com — published by Hartzer.it.com